Predictive factors of central lymph node metastasis in papillary thyroid carcinoma.

PURPOSE: The aim of this study was to evaluate the correlation between central lymph node (CLN) metastasis and clinicopathologic characteristics of papillary thyroid cancer (PTC). In addition, we investigated the incidence and risk factors for contralateral CLN metastasis in unilateral PTC. This study suggests the appropriate surgical extent for CLN dissection.
METHODS: A prospective study of 500 patients with PTC who underwent total thyroidectomy and prophylactic bilateral CLN dissection was conducted.
RESULTS: Of 500 patients, 255 had CLN metastases. The rate of CLN metastasis was considerably higher in cases of younger patients (<45 years old) (P < 0.001; odds ratio [OR], 2.357) and of a maximal tumor size greater than 1 cm (P < 0.001; OR, 3.165). Ipsilateral CLN metastasis was detected in 83.1% of cases (133/160) of unilateral PTC, only contralateral CLN metastases in 3.7% of cases (6/160), and bilateral CLN metastases in 13.1% of cases (21/160). The rate of contralateral CLN metastasis was considerably higher in cases of PTC with a large tumor size (≥1 cm) (P = 0.019; OR, 4.440) and with ipsilateral CLN metastasis (P = 0.047; OR, 2.613).
CONCLUSION: Younger age (<45 years old) and maximal tumor size greater than 1 cm were independent risk factors for CLN metastasis. Maximal tumor size greater than 1 cm and presence of ipsilateral CLN macrometastasis were independent risk factors for contralateral CLN metastasis. Therefore, both CLN dissections should be considered for unilateral PTC with a maximal tumor size greater than 1 cm or presence of ipsilateral CLN macrometastasis.

INTRODUCTION

Thyroid carcinoma accounts for approximately 1% of all tumors and one third of all head and neck tumors. Papillary thyroid carcinoma (PTC) makes up 85% to 90% of all thyroid carcinoma cases, with a reported 10-year survival rate of more than 90% [1]. Although PTC has a good prognosis, central lymph node (CLN) metastases are common. The occurrence of CLN metastasis ranges from 40% to 90% of PTC cases [2]. CLN metastasis is a well-known independent risk factor in local recurrence [3,4,5] but has little adverse effect on survival [5,6]. However, this conventional knowledge has become debated due to a large, recent case-based study which showed an increased mortality rate in patients with regional lymph node metastasis [1].

Prophylactic central lymph node dissection (CLND) is an accepted procedure for patients with PTC. However, CLND might increase the rates of hypoparathyroidism and vocal cord palsy, especially when bilateral CLND is performed [7]. Thus, identification of predictive factors for CLN metastasis might prevent unnecessary contralateral CLND.

Recently, PTC has been diagnosed at earlier stages through screening. When prophylactic CLND is performed, many CLN metastases are microscopic deposits [8].

In this study, we examine the clinicopathologic characteristics of PTC with CLN metastasis and investigate the incidence and risk factors for contralateral CLN metastasis in unilateral PTC. In addition, we describe the incidence and characteristics of CLN micrometastases and suggest the appropriate surgical extent for CLND.

METHODS

Eligibility

Our prospective study was approved by the Institutional Review Board of Chungnam National University Hospital, and written informed consent was obtained from participants. In 2001 and 2012, 500 patients with PTC who underwent a total thyroidectomy with bilateral prophylactic CLND were enrolled in this prospective nonrandomized study. Patients with clinically node-negative necks were evaluated preoperatively through neck ultrasonography, fine needle aspiration biopsy, 18F fluorodeoxyglucose (18F-FDG) PET/CT, and physical examination. All patients were clinically node negative. Being clinically node-negative was defined as having no signs of enlarged lymph nodes (LN) during preoperative ultrasonography and no glucose uptake in LN during preoperative 18F-FDG PET/CT.

Surgical procedure

All patients underwent total thyroidectomy and prophylactic bilateral CLND. First, the entirety of both lobes of the thyroid gland was removed, and then bilateral CLND was performed. CLNs are group VI LNs that surround both lobes of the thyroid and include the pretracheal, paratracheal, and prelaryngeal LNs within an area bordered laterally by the carotid sheath, at the top by the hyoid bone, at the bottom by the sternal notch, and dorsally by the prevertebral fascia. The thymus was preserved through separation from the CLNs, and the parathyroids and recurrent laryngeal nerves were also identified and preserved. Parathyroid glands that could not be preserved in situ were autotransplanted into the ipsilateral sternocleidomastoid muscle. At the end of the operation, the surgeon cut separately into the pretracheal nodes (immediately anterior to the trachea) and the paratracheal nodes (adjacent to the trachea on either side), categorized as described above, and sent them to the pathology department for examination. In unilateral thyroid cancer, ipsilateral CLNs are pretracheal and ipsilateral paratracheal LNs, while contralateral CLNs are only contralateral paratracheal LNs.

Statistical analysis

Statistical analysis was performed using SPSS ver. 12.0 (SPSS Inc., Chicago, IL, USA). Pearson chi-square or Fisher exact test for categorical variables was used in the univariate analyses of the clinical characteristics. Variables with a P-value of less than 0.05 in the univariate analysis were included in multivariate logistic regression analysis.

RESULTS

The clinicopathologic characteristics are presented in Table 1. Of the 500 patients, 255 had metastatic CLNs. Univariate analysis suggested that younger age (<45 years old), male sex, a large tumor size (≥1 cm), bilateral multiple nodules, and the presence of capsular invasion or extrathyroidal extension were significant factors in CLN metastasis (Table 1). With CLNs removed in the multivariate analysis, the rate of CLN metastasis was considerably higher in the cases of younger patients (<45 years old) (P < 0.001; odds ratio [OR], 2.357) and large tumor size (≥ 1 cm) (P < 0.001; OR, 3.165) (Table 2).

Table 1

Correlation between clinicopathological characteristics and central lymph node metastasis in papillary thyroid cancer

Values are presented as number (%) or mean ± standard deviation. CLN, central lymph node.

a)Pearson chi-square test for categorical variables was used for univariate analyses of the clinical characteristics.

Table 2

Correlation between clinicopathologic characteristics and central lymph node metastasis in papillary thyroid cancer by multivariate logistic regression analysis

SE, standard error; CI, confidence interval.

The trend of CLN metastasis in unilateral PTC patients is depicted in Fig. 1. Of 341 patients with unilateral PTC, 160 had CLN metastases. We excluded isthmic PTC in patients with unilateral PTC analysis.

Fig. 1

Trend of central lymph node (CLN) metastasis in unilateral papillary thyroid cancer (PTC). CLND, central lymph node dissection.

Ipsilateral CLN metastasis was detected in 83.1% of cases (133/160) of CLN metastases in patients with unilateral PTC, only contralateral CLN metastases in 3.7% (6/160), and bilateral CLN metastases in 13.1% (21/160). The clinicopathologic characteristics of patients with contralateral CLN metastases are shown in Table 3. In the univariate analysis, large tumor size (≥1 cm), lymphovascular invasion, and positive ipsilateral CLNs were significant predictors of contralateral CLN metastases in unilateral PTC (P < 0.05) (Table 3). In the multivariate analysis, the rate of contralateral CLN metastasis was considerably higher in the cases of PTC with a large tumor size (≥1 cm) (P = 0.019; OR, 4.440) or ipsilateral CLN metastasis (P = 0.047; OR, 2.613) (Table 4).

Table 3

Correlation between clinicopathological characteristics and contralateral central lymph node metastasis in unilateral papillary thyroid cancer

CLN, central lymph node; LN, lymph node.

a)Pearson chi-square test or Fisher exact test for categorical variables was used for univariate analyses.

Table 4

Correlation between clinicopathologic characteristics and contralateral central lymph node metastasis in unilateral papillary thyroid cancer by multivariate logistic regression analysis

SE, standard error; CI, confidence interval.

Permanent hypocalcemia occurred in 22 patients (4.4%), while permanent recurrent laryngeal nerve paralysis did not occur in the present study (Table 5).

Table 5

Complications (n = 400)

Eighty-nine patients had CLN micrometastases (Table 6). The rate of CLN micrometastases was higher in cases with a maximal tumor size of less than 1 cm (P = 0.002). All the CLN micrometastases occurred on the ipsilateral side of the tumor.

Table 6

Characteristics of subgroups of nodal involvement in papillay thyroid carcinoma

Values are presented as number (%).

LN, lymph node.

a)Pearson chi-square test for categorical variables was used for univariate analyses of the clinical characteristics.

DISCUSSION

PTC responds relatively well to surgery and has a good prognosis. However, PTC might recur if the first operation is conducted improperly. A standard treatment for PTC has not been established. A growing number of studies present prophylactic CLND in synchronization, and some authors even consider CLNs as important as the primary tumor [9]. Therapeutic CLND should be performed on patients with PTC who have preoperative clinical LN metastasis, despite the controversy over clinical N0 (cN0) PTC patients [7,10,11]. Most studies agree that CLNs are involved most frequently in PTC with LN metastasis. The LNs in the central group are most commonly involved in metastases, so there might be a significant risk of recurrence in the central group [12,13]. In addition, reoperation is relatively difficult and complicated in patients who develop loco-regional recurrence in the central group [14,15]. Several studies suggest that prophylactic CLND should not be performed routinely for the majority of cN0 PTC patients, even though a greater incidence of subclinical LN metastasis is expected [16]. Performing CLND has significant associated morbidities, such as recurrent laryngeal nerve injury and hypoparathyroidism. Therefore, it is important to determine the surgical extent of prophylactic CLND in the initial operation for patients with PTC. Recent studies on CLND report transient vocal cord palsy in 3% to 6% of cases and permanent hypoparathyroidism in 3% to 4% of cases [17,18,19,20,21]. In this study, half the patients developed temporary hypoparathyroidism, and 4% continued to experience symptoms for more than 6 months after surgery. These results are similar to those in previous research [7,22]. Permanent recurrent laryngeal nerve paralysis did not occur in the present study. Careful preoperative nodal evaluation for metastasis should be done in order to avoid unnecessary CLND for pN0 patients. However, CLNs are usually small and difficult to identify with ultrasound. Recent studies have shown that the sensitivity in identifying CLN metastases using ultrasonography is 40% to 60% [23,24].

Various predictive factors and staging systems have been used to determine PTC prognosis. Age, gender, tumor size, lymphovascular invasion, extrathyroidal extension, and LN metastasis are generally accepted factors in survival and local recurrence. LN metastases are an independent risk factor for loco-regional recurrence and distant metastasis [25]. Evidence from large population-based studies shows that patients with regional LN metastases have poor prognosis and higher mortality [1,17]. Numerous studies have found CLN metastases in 50% to 70% of PTC patients [13,20,26,27]. In the present study, 51% of PTC patients (255/500) had CLN metastases. Most studies agree that ipsilateral CLN metastases are frequently involved in the LN metastasis of PTC [8,13,28]. Prophylactic CLND ipsilateral to the tumor associated with total thyroidectomy could be an effective strategy for reducing the rates of permanent hypoparathyroidism and hoarseness. Generally, for unilateral PTC, the rate of metastasis to contralateral CLN was very low, and the rate of disease-free survival did not improve after bilateral CLND, indicating that contralateral CLND is unnecessary [12,29]. Among the unilateral PTC patients with CLN metastases participating in the present study, contralateral CLN metastasis occurred mostly in the patients who had ipsilateral CLN macrometastasis or a maximal tumor size greater than 1 cm (Tables 4, 6).

We recommend that total thyroidectomy with simultaneous completion of contralateral CLND be performed in the presence of ipsilateral CLN macrometastasis in intraoperative frozensection pathology. This approach could limit unnecessary contralateral CLND for node-negative patients based on intraoperative pathological findings. According to the results of this study, both prophylactic CLND should be considered for unilateral PTC with a large tumor size (≥1 cm) or the presence of ipsilateral CLN macrometastasis. In contrast, prophylactic ipsilateral CLND has a satisfactory surgical extent for the initial operation in patients with PTC who have a small tumor and no ipsilateral CLN macrometastasis (Fig. 2).

Fig. 2

Suggested scheme of decision making in unilateral papillary thyroid cancer (PTC). The ipsilateral central lymph node (CLN) metastasis could be evaluated by clinical examination or intraoperative pathology. CLND, CLN dissection.

There were no local recurrences during the follow-up period for this study. Local recurrence and disease-free survival could not be described because of the short follow-up duration. A long-term follow-up of the present study is necessary to evaluate local recurrence and disease-free survival.